# DEPRECATED (2026-10-28): single-step ball reflow
> **Deprecated.** Two-step reflow (thin flat pre-tin, then PnP into tacky flux + supplemental Ø0.30 balls) is the OSSO process. See docs/process_reflow.md. This file is kept for history only.

Supersedes the two-reflow "pre-tin then place" flow of drawings/osso_process_flow_v1 (steps 12-19) as the DEFAULT. Pre-tinning stays as the FALLBACK.
Drawing: drawings/versions/single_step_reflow_pad_guidance_older/2026_10_28_single_step_reflow_pad_guidance_v3.png (v2 and v1 superseded 2026-10-28). Ball-count table: project/runs/q_run_20261028/item5_reflow/ball_count_table_v3.json (script make_reflow_v3.py).

## Steps (after laser silk/mask/copper isolation and cleaning)
1. **Flux patch.** One tacky gel flux patch per terminal. It covers the terminal landing area AND the ball spot(s) as one continuous patch, so the molten ball has a fluxed path onto the terminal.
2. **Ball drop.** 0.30 mm SAC305 balls (one ball size for every pad) from the N1 feeder (puff seat) onto the ball spots. Ball centre within about one ball diameter (≤ 0.30 mm) of the terminal edge, on the exposed pad toe or side; never under the terminal.
3. **Camera check.** Every spot has its ball count; extra balls removed, missing ones re-dropped.
4. **Place parts** into the same flux. The tacky flux holds both the balls and the parts.
5. **One reflow.** Bed soak profile as process_v3 step 10 (large pour = heat sink). Balls melt, wet the pad along the flux, and wick to the terminal.
6. IPA clean, inspect, test (process_v3 steps 11-12).
No-populate pads: flux + balls, reflowed in the same pass (keeps the "no bare copper" rule of Oct 16).

## Ball count and placement rules (v3, 2026-10-28: one ball size, Ø0.30, for all pads)
- One ball size for every pad: Ø0.30 SAC305, volume 0.01414 mm³ (1.73x a Ø0.25 ball). No row 2 anywhere.
- Volume target = pad area x 0.06 mm (≈ a 0.12 mm stencil print at ~50 % metal). n = ceil(V/Vb) for 0603 and larger, limited to rows 0 and 1.
- ONE flux patch per terminal, covering the terminal landing and all ball spots.
- Row 0 touches the terminal edge (gap ~0.02). Row 1 sits directly behind it (near ball edge ~1.07 D = 0.32 mm from the terminal). Gull-wing leads (SOT, SOIC) use the toe, then the heel side.
- 0402 and fine pitch (≤ 0.5 mm, QFN/QFP): the minimum, ONE ball per terminal, centred on the toe (bridging risk). QFN: the Ø0.30 ball is wider than the 0.25 pad by 0.025 per side; 0.225 mm remains to the neighbour pad edge (rule ≥ 0.15).
- Two-terminal parts: SAME ball count and placement on both terminals (asymmetric melt timing causes tombstoning).

| Pad | needed for 100 % | used | layout (per terminal) | % of paste-equiv. volume |
|---|---|---|---|---|
| 0402 (0.50x0.60) | 2 | 1 | toe row0: 1 | 79 |
| 0603 (0.80x0.90) | 4 | 4 | toe row0: 2; toe row1: 2 | 131 |
| 0805 (1.0x1.3) | 6 | 6 | toe row0: 3; toe row1: 3 | 109 |
| 1206 (1.6x1.6) | 11 | 10 | toe row0: 5; toe row1: 5 | 92 |
| SOT-23 lead (0.6x1.0) | 3 | 3 | toe row0: 2; toe row1: 1 | 118 |
| QFN pad 0.5p (0.25x0.7) | 1 | 1 | toe row0: 1 | 135 |
| SOIC lead (0.6x1.6) | 5 | 5 | toe row0: 2; toe row1: 2; heel row0: 1 | 123 |

## Fallback: pre-tinning
Use when wetting is poor (coupon test or AOI: ball not merged with terminal, dewetting on laser-opened copper): flux, balls, reflow the bare board, camera check, flux again, place, second reflow (the old v1 flow). Decide per board, not per joint.

## FLAG (physics)
v3 (single Ø0.30 size) removes row 2: every ball is within ~1 D of the terminal edge. Remaining:
- 1206 reaches 92 % (10 balls, rows 0+1 full at 1.6 mm pad width); an 11th ball would need a row 2. Accepted as ~100 %; check fillet in the coupon test.
- 0402 is 79 % with its single ball (minimum without bridging; 2 balls would be 157 % on a 0.5 mm pad).
- 0603 row 1 overhangs the pad end by ~0.13 mm (0.62 needed vs 0.495 exposed toe); SOT-23 by ~0.07, QFN by ~0.04. The balls wet back onto the pad inside the flux patch; check in the coupon test.

## Test before adopting (proposed TODO 17 replacement)
17. [ ] Single-step coupon test: flux patches + Ø0.30 balls (per the v3 table) + 0402/0603/0805/SOT-23 parts on the same flux, one reflow. Cross-check wetting on ENIG and on laser-opened copper, ball merge rate, tombstones, bridging at 0.5 pitch, and fillet height. If wetting on laser-opened copper is poor, run the pre-tin fallback on half the coupon for comparison.
